By Johnny Walker.
In the previous post, I shared my students’ projects and reflections about working with AI to research and create images of murals that represented different nations’ migration journeys to our city. Below are my reflections and new learnings about this project and guiding students to inquiry about AI.
The Drafting & Reflection Process
- When students make a record of multiple rounds of questioning and research, they gain awareness of their own thinking and understanding. Students document each round of questioning, research, and give the AI re-direction to build a time-lapse of how they learned about a topic. When they reflect on that learning process and see how the language of their prompt increased in specificity and sophistication, they can track and gauge how they learned.
- When students guide A.I. in multiple drafts, they act as an executive who delegates, rather than as a passive consumer. And, in directing an AI assistant, students build awareness of how their ability to manage others, to clearly articulate a vision, to give actionable feedback, to engage in questioning as a means to learn, to corroborate evidence, to be a final arbiter of taste, and be ultimately responsible for the final product, is vital when they (not their AI assistant) are held accountable for the final product.
- When students shared and listened to reflections on the positives and negatives of working with AI, they ultimately constructed greater understanding. As they made connections and articulated appreciations for each other’s reflections, students saw the value of sharing as a community to find answers and build additional understandings.
Evaluating Sources & Evidence
- Students should always check the sources from which the AI has based its conclusions. They are then able to make evaluations and determine the extent to which they should also choose to prioritize one piece of evidence over another.
- Questioning protects students against manipulation. It is easy to be swayed by the ‘Wow Factor’ of AI image-generation. While we are amazed and inspired by AI images that respond to our prompts, once we apply scrutiny (observing details, questioning and research), the AI’s considerable flaws are exposed. Carefully vetting a source of evidence and corroborating becomes a valuable skill. It is a powerful learning experience to not trust our initial emotional or sensory reactions to a stimulus. We need to guard against appeals to our sense of wonder as being potentially manipulative and guiding us toward ignorance, and instead look to our own skills of questioning, gathering reliable evidence and reason to guide a process to avoid being bamboozled.
- Using multiple AI platforms in the same project allows students to guard against AI validation that could reinforce false ideas. For example, using Google AI to research and find original source material for questions related to images generated by another AI, Chat GPT, keeps the student from getting potentially lost or stuck in a feedback loop or delusional spiral that can result from relying on one AI platform as a source. This will allow students to shape the direction of their research and corroborate evidence rather than it being shaped by a single AI source.
- Modeling questions that involve “sourcing” will increase student learning outcomes in future applications. When students see how asking questions related to the origin of the source and its reliability can improve their ability to spot misinformation and thus gain appreciation of asking such questions.
Fostering Curiosity
- Students, during a questioning process, maintain a state of healthy skepticism that leads to more specific research.
- When students generate their own questions as the first step of a research or creative process, they ground their learning in a powerful resource: their own curiosity. Self-generated questioning further allows students to suspend judgement, to gradually piece together distinct facts, to eventually connect facts together to create greater understanding, to spot misinformation, correct misconceptions, and to intentionally shape the direction of a project.
- When students complete a project with AI, concluding with lingering questions can establish a sophisticated habit. Questions that students pose at the end of a particular project can ground their future learning as a process that accepts uncertainty, and views learning as a continual journey of inquiry. And, when students finish a project with generating questions, you can assess the extent to which their questions reflect a sophistication about what they have already learned.
Enabling Accessibility
- AI Speeds Research. Rather than taking weeks, a student can spend minutes or hours engaging in a cycle of questioning, collecting evidence, making connections and then revising their new guiding directions multiple times. Each journey through the cycle spirals their learning to build a layered and more nuanced understanding. Notwithstanding the limitations of its reliability, when properly used, AI nevertheless streamlines and speeds a process for humans to drive and shape their research and creativity. Teachers have greater opportunities for jig-sawing content and allowing for student choice when the process, not the product, is what is assessed.
- Iterating prompts into an image generator is accessible and applicable to multiple subjects. Whether students are creating a mural about the Asian migration to California in a World History class in Los Angeles, designing a biome to house a tropical climate zone in Biology, creating a book cover for a novel in English, or designing a roller coaster in Math, AI images provide an accessible point of entry for all students. Their process of questioning, research and rewriting more specific prompts to iterate a product that they have greater ownership and understanding over, can be universally adapted.
Questioning the Use of A.I.
- It takes a leap of faith and vulnerability on the teacher’s part to bring AI to the classroom. However, it also reinforces confidence in the important work that we do to create environments and challenges that allow for uncertain outcomes, yet maintain guardrails and encourage communication. It is exciting to employ our students as valuable sources of analysis and insight, and give them opportunities to drive their own, influence each others’ learning, and teach us too.
- Student resistance to working with AI is also an opportunity for the entire class to grapple with the moral and ethical implications of its use. Students may have concerns related to AI’s unsustainable environmental effects, deleterious energy or water use, copyright infringement, unregulated privacy and personal data use, deleterious mental health effects, dissemination of misinformation, potential to damage social cohesion, increasing loneliness or isolation, enflaming political extremism, to misappropriate or exploit culture, to threaten the existence of artistic, musical or analytical careers, to create false historical narratives, exacerbate economic inequality…. Take your pick! There are so many disturbing and destabilizing issues directly related to AI that our students may already be aware of or may want to discuss. One of my student’s refusal to use an AI artist gave her the opportunity to create her own artwork and share why she did with her classmates, resulting in a powerful learning experience for us all. A short project using it can open opportunities to debate, to discuss, or to create related action projects.
- AI should not be used in the shadows. Transparency is vital to expose its dangers, its strengths, and prevent misuse. As “sunlight is the best disinfectant”, intentionally bringing our students’ AI use into the classroom, sharing our collective experiences with it, and exposing its strengths and flaws allows students to view it not as a panacea to outsource their own critical thinking, or as a forbidden place to secretly visit. When young people (and teachers) use the technology together and in the open, it creates spaces to explore as a community, to experiment with its professional applications, and to temper its dangers with adult guidance.
In seeing the work and process that our students engaged in, we realized that while A.I. is limited, the human potential for creativity and curiosity is not. And, when we use our own curiosity and natural intelligence to guide us rather than being guided by artificial intelligence, we can arrive at greater knowledge of ourselves, our history, our diverse communities, the places throughout the world that influence us, and connect more deeply with each other.
Johnny Walker teaches 8th Grade World History at Campbell Hall Episcopal School in Studio City, California. Prior to that, he taught high school history for ten years at Triumph Charter High School in Los Angeles. Johnny also works with The Right Question Institute, teaching educators to bring the Question Formulation Technique into their practice and develop classroom communities of inquiry, curiosity, and critical thinking. Johnny holds a B.A. in History and Political Science from Yale University and an M.A. in Teaching Secondary Social Science from The University of Southern California.
